Talk with other people in your field (real people, not cherry-picked histories that could be made up or at least heavily colorized), and not only the ones that

Talk with other people in your field (real people, not cherry-picked histories that could be made up or at least heavily colorized), and not only the ones that talk at conferences. You'll realize that your peers know much less and are much less capable than you attribute to them. > And I read online how person X just brushed over documentation and got things running solid, They didn't. It took them a significant amount of time, it's just very hard to record and later describe this process of trial, error, and backtracking. It's much, much easier to polish the procedure for final publication. > came to hackaton and developed out of thin air what startups with teams release crappy versions of in several months. This "thin air" took significant amount of preparation. The hackaton product is bug-ridden, resource-hungry proof of concept that can't serve more than a handful users, crashes badly on slightly invalid data, and cannot be maintained in production (data backups? recovery after server crash? user management? logging? status monitoring?). Not to mention that topics for hackatons are chosen specifically to suit well the form of making a PoC in two days. Useful things usually do not fit this time frame.